Yi Technology -- Yi-M1

[kazanunda]

I've been using Yi M1 for about a year now and I can agree that it is a really nice camera for manual lenses.

Though it does have it's quirks. For example sometimes if I shoot it in portrait orientation, the bottom side of the image will get darkened as if a gradient was pasted over it (bottom side of the sensor when the camera is in landscape orientation). The weirdest part is that it's inconsistent, it doesnt happen all the time. I have yet to test it with different lenses and the same lenses on the spare Yi, just curious if you have had the same issue?

 

[VidThreeNorth]

I haven't seen anything like that happen on my camera. If you can upload an example or two I might have an idea about it.

 

[kazanunda]

Spoiler: image

I have a couple other examples of it, but I cant seem to find them at the moment. The left side is heavily darkened.

 

[VidThreeNorth]

That looks like some kind of shutter failure. Unfortunately, I have no detailed information about how these new shutters work. Is this happening during rapid sequence "burst" shooting or is it happening during single shots with a time gap?

 

[kazanunda]

During single shots with a time gap, I rarely do bursts because of it's limited buffer. But wouldn't shutter sticking leave a straight break line where it got stuck passing over the sensor, instead of a gradient?

Edit I have tested the spare unused M1 today with the 28mm kamlan. I suspect it may be lens related as the images showed the exact same gradient. Curiously, it only happens in portrait mode and the bottom side is darkened regardless of how I rotate it in the portrait mode. I couldnt find this issue on portrait shots with the kamlan 50mm f1.1 mark2, so at this point I am fairly certain its lens related.

 

[VidThreeNorth]

I have no real information about how modern shutters work, so I was just playing with an idea. From what descriptions I have seen, it seems like they are powered by variations on "linear motor" technology which is similar to the top quality lens focussing drives. It would take a surge of power to get the blades up to speed, and that would take a capacitor of some kind. I was thinking that the first shutter curtain capacitor might not be charging fully, so that it might slow down by the end of the run.

If the second blade is moving faster then it could catch up to the first blade, reducing the exposure.

I don't know that lens. Is it M43 mount? If the aperture is actually being opened or closed by the camera, then the aperture might simply be too slow. Are other lenses working properly?

 

[kazanunda]

Other lenses are working fine.

Kamlan only makes fully manual lenses.

Recently got a 180mm f2.8 zeiss, so that will be interesting on MFT, just need to wait for the adapter to arrive.

 

[VidThreeNorth]

Since it is a fully manual lens then neither the shutter nor the aperture should have any such problem. I can't see any answer to it. Since all the other lenses are working, then my "absolutely brilliant" suggestion is: "just don't use that lens."

 

[mmsumulong]

I just picked it up for about 200 with both lenses. This is my first step pasted a point and shoot and I am loving it.
. . .

It looks like you're doing well so far. As I wrote above, the 42.5mm is a very good lens. Unfortunately, I have never seen batteries sold separately for this camera. If you ever see one than let us know.

https://www.amazon.com/gp/product/B07VNRVX6H/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1

I purchased this myself and is slightly better than the OEM battery. Do note that this is a cheap camera but workable as a secondary/backup camera or something to toy around for experienced users. But is still fine for novices. There are obviously better cameras than this. But the batteries work great. And the charger is a plus! [edited]

 

[kazanunda]

Regarding obtaining batteries in europe, I can recommend electropapa from germany: Li-Ion-battery - 900mAh (8.8V) - for camera, digicam, DSLR replaces XiaoYi BXM-10 - Electropapa

To be honest they arent as good as original batteries, but they work well enough. Biggest issue in getting them from china is that they arent being sold with chargers at all. And solo batteries cant be shipped by air.

I also heard some people recommending kodak batteries because of similar shape and size. The indents are in the wrong corners and the terminals on the battery are reversed, so it wont work.

 

[VidThreeNorth]

Much thanks guys! Unfortunately, the Amazon vendor linked will not ship to Canada, so I will have to check further. Maybe the Electropapa site will. I'll look into it later.

 

[mmsumulong]

Much thanks guys! Unfortunately, the Amazon vendor linked will not ship to Canada, so I will have to check further. Maybe the Electropapa site will. I'll look into it later.

The ebay seller(s) ship(s) worldwide. But loses out on Amazon. But again I have purchased these on Amazon and they are rated better than the OEM and the bundled charger is a boon too.

DSTE 3PCS BXM-10 BXM10 Rechargeable Li-ion Battery for XIAOYI M1 | eBay

DSTE US Plug Battery Charger with Car Adapter for XiaoYi BXM-10 Camera Battery | eBay

 

[VidThreeNorth]

Much thanks again. I ordered the charger set from the Amazon link in #69 and it arrived a couple of weeks ago. The batteries and charger are all working. It was all "just in time." My original battery was getting thick, meaning that the internal battery was deteriorating and not safe. I have recently bought a Panasonic G85 which will become my main "4K" video camera for now. That means I will mainly be using the Yi-M1 for still pictures in the future. Those, along with my Panasonic GF3 form the core of my M43 system.

The following link points to the topic for my "Autumn" video. This will probably be the last video that post that came completely from the Yi-M1. I was particularly happy with how the colours turned out in it.

"20101020 Toronto Autumn"
20101020 Toronto, Autumn

 

[VidThreeNorth]

I have finally gotten around to buying a standard colour card. I bought a Datacolor "SpyderCheckr 24". I already had an Elvid slate which has a partial set of colours. Combining the two gives me a good range. Apparently, Datacolor also has a free downloadable program to use with their colour card. I had a method of evaluating colours in mind, but I will probably download and install that program as well.

I spent most of today recording videos for evaluation from my Yi-M1 and my Panasonic G85. These are frames captured from some of those videos.

Conditions: This was a morning file set (around 9:30-ish) on a clear day. I was careful not to locate the card in shadow, but due to the angle, the card did reflect light more directly towards the Panasonic. The cameras were located on my 3D bar and were about 3' away from the card. The card is sitting on top of the Elvid slate to keep it off the grass. The lens on the Yi-M1 is my Panasonic 12-42 F3.5-5.6 and the lens on the G85 is the 12-60 F3.5-5.6. Focal lengths were around 30-35mm. Exposure settings were not recorded. Both cameras were set to Auto exposure and Auto white balance. ISO should have been 200 for both. Exposure compensations are noted for each clip. I tested EV = +0.0 for each of the profiles that I was interested in, and for some I also tested EV = -1.0 and sometimes also EV = +1.0, all according to my current interests. I am not uploading captures from all the files I made. These are just the ones I think are most interesting.

NOTE: I might post evaluations later or I might not. There are "how-to" videos and web pages that teach evaluation methods, so anyone can do their own. You don't need my comments.

[2020-04-26 15:13]
The Panasonic G85 files have been uploaded to "G85 w/12-60mm F3.5-5.6 lens"

[All JPEGs compressed level 1]
"YiM1-P4250002-Standard-ev0_0-12h42m50s970.jpg"
"YiM1-P4250005-Portrait-ev0_0-12h55m21s883.jpg"
"YiM1-P4250006-Portrait-evMinus1_0-13h09m23s876.jpg"
"YiM1-P4250009-Vivid-ev0_0-15h00m51s788.jpg"

 

[VidThreeNorth]

YiM1, Panasonic 12-42 F3.5-5.6 zoom (version 1)

Before I got the colour checking card, I did not have a really accurate way of evaluating how my cameras were treating exposure. As time went on, I was developing a general idea, but it was a slow process of trial and error. My previous conclusion was that the contrast and "dynamic range" of the three colour profiles of the Yi-M1 were very close. I did notice that the "Portrait" seemed to give a bit more shadow detail, so I thought it probably had a bit more dynamic range. I had little experience at all with the "Vivid" profile because it seemed too saturated for what I was recording. Now that I have the colour checking card, the first thing I wanted to find out was what differences if any there were.

I used the pixel reading ability of Corel Paint Shop Pro to read five pixels in each of the six grey squares of the captures ("Standard" w/EV = +0.0 and "Portrait" w/EV = +0.0". For convenience, I will call these colours (in order from darkest to lightest) "Black", "Black + 1", "Black + 2", "White - 2", "White - 1" and "White".


Graph Data:

The following are the brightness "[luma"] values in 8-bit form for some of the sample images:

[All JPEGs compressed level 1]

"YiM1-P4250002-Standard-ev0_0-12h42m50s970.jpg"

"Black" = 17.5 (below 20)
"Black + 1" = 60.9
"Black + 2" = 130.2
"White - 2" = 184.9
"White - 1" = 223.8
"White" = 246.0

"YiM1-P4250005-Portrait-ev0_0-12h55m21s883.jpg"

"Black" = 30.8
"Black + 1" = 78.2
"Black + 2" = 144.2
"White - 2" = 198.3
"White - 1" = 232.0
"White" = 254.4 (clipped)


"YiM1-P4250009-Vivid-ev0_0-15h00m51s788"

"Black" = 13.6 (below 20)
"Black + 1" = 59.6
"Black + 2" = 132.6
"White - 2" = 193.6
"White - 1" = 234.1
"White" = 254.8 (clipped)


"G85-00010-Standard-ev0_0-13h26m54s730"

"Black" = 54.9
"Black + 1" = 104.5
"Black + 2" = 163.0
"White - 2" = 207.3
"White - 1" = 235.3
"White" = 253.5 (clipped)

[2020-05-09 19:41 reformatted the above for consistency]

Special Cases:

There are three "special cases" I should explain. All have to do with the "White" or "Black" values. The far limit values are "0" and "255" If even 1 of the values of red, green or blue component values for any of the pixels tested is equal to either "0" or "255", then I have marked the resulting Luma value as "clipped". Since I only tested five pixels for each colour, it can be assumed that more pixels had such values. For such a case, I would prefer to adjust my exposure to better "contain" the exposure of that brightness level if it can be practically done. So if I see clipping, then I will probably adjust to eliminate it.

The "under 20" values are not as urgent. "Under 20" is a broadcast standard for "black" That is to say that any value under "20" can be presumed to be "black" for the purposes of North American television. That is all it means. It is commonly called "legal black", but that is a misconception. There is no "law" being violated. However, values below 20 can be "noisier" and are thus less reliable. But there is no guarantee that a value higher than 20 will be free of noise. One needs to study one's camera(s) to know how much of "under 20" can be recovered in post.

The Graph:

I graphed some of the values as "reciprocity curves" and there are a few things that can be said about the results.

First, the contrast situation was a bit worst than I thought -- not much, but definiitely worse. The reciprocity curves for the Yi-M1 are almost identical. Yes, there is slightly less contrast in "Portrait" profile, but only around a half stop less than the "Standard" profile. The exposure of of the "Portrait" profile appears to be about 1/3 stop lighter than "Standard". This accounts for part of the increased shadow detail. Unfortunately, to make use of the full dynamic range in "Portrait", I would need to expose at at about "EV = -1/3", which gives up some of that shadow detail.

In the "Standard" profile, the "Black" sample brightness is entirely below 20, which is "legal black". It was not clipped, but recovery is unreliable. I have recovered shadows in some cases, but my experience confirms this limitation.

The contrast for "Vivid" is even worse than "Standard". I did not bother to graph it. Now that I know how limited its dynamic range is, I don't intend to use it again.

The "G85 Standard" profile is included as a control (for comparison).

Finally, to find a single number value to compare the reciprocity for each profile I subtracted the "black + 1" value from the "white - 1" value. For these values, the lower results indicate less contrast. It is not necessarily true that one always wants less contrast, but for most of what I do, that would be my general goal.

In fact, I probably will not even use the G85 "Standard" profile very much. The G85 has profiles with even less contrast, which I will probably use more often. I doubt if I will use the Yi-M1 for video much in the future, but if I do, from now on, I will probably only use the "Portrait" profile. Since I use the "Histogram" function on that camera when chosing my exposure, I generally set it to "protect" my highlights that way. If I didn't, then I would probably try to reduce the exposure by "EV = -1/3" from now on.

Yi-M1 Standard 223.8 - 60.9 = 162.9
Yi-M1 Portrait 232.0 - 78.2 = 153.8
Yi-M1 Vivid 234.1 - 59.6 = 174.5
G85 Standard 235.3 - 104.5 = 130.8